Defect diagnosis

Abstract

A method for diagnosing a defect is provided. A first candidate pair comprises a first defect candidate and a second defect candidate. A first pattern is generated to distinguish one or more faults of the first defect candidate from one or more faults of the second defect candidate. The first defect candidate is removed responsive to determining that the first pattern does not detect the first defect candidate and determining that an automatic test equipment (ATE) failure log associates the first pattern with failure. Removing the first candidate pair, as well as additional candidate pairs when possible, promotes diagnosis efficiency by reducing a number of computations required.

What is claimed is: 
     
       1. A method, comprising:
 receiving a list of one or more defect candidates; 
 determining a plurality of candidate pairs from the list to establish a set of candidate pairs, wherein the determining comprises:
 generating a first candidate pair comprising a first defect candidate and a second defect candidate responsive to determining that the first defect candidate and the second defect candidate have a same defect; and 
 generating a second candidate pair comprising a third defect candidate and a fourth defect candidate responsive to determining that the third defect candidate and the fourth defect candidate have a same fault effect; and 
 
 at least one of:
 dropping the second candidate pair from the set of candidate pairs to generate a reduced set of candidate pairs responsive to determining that the third defect candidate and the fourth defect candidate have the same fault effect; 
 performing candidate pair grouping, comprising:
 distinguishing a fifth defect candidate of a third candidate pair in the reduced set of candidate pairs from a sixth defect candidate of the third candidate pair using fault inactivation, wherein the fault inactivation comprises forming a first pattern that detects the fifth defect candidate but not the sixth defect candidate; 
 grouping the third candidate pair with a fourth candidate pair in the reduced set of candidate pairs to create a first group responsive to determining that a first portion of a first circuit associated with the third candidate pair does not overlap a second portion of the first circuit associated with the fourth candidate pair; 
 inserting the first group into a second circuit; and 
 forming a second pattern for at least some of the first group based upon the second circuit; or 
 
 performing sub-circuit analysis, comprising:
 identifying a convergent point of a third circuit on a first portion of the third circuit that is associated with a seventh defect candidate of a fifth candidate pair and a second portion of the third circuit that is associated with an eighth defect candidate of the fifth candidate pair, wherein a third portion of the third circuit that is associated with the convergent point is merely able to be affected by at least one of the seventh defect candidate or the eighth defect candidate via the convergent point; 
 removing at least some of the third circuit after the convergent point to generate a reduced circuit; and 
 determining whether the seventh defect candidate is equivalent to the eighth defect candidate based upon the reduced circuit. 
 
 
 
     
     
       2. The method of  claim 1 , wherein the second circuit and the third circuit are a same circuit. 
     
     
       3. The method of  claim 1 , wherein the second circuit is different than the third circuit. 
     
     
       4. The method of  claim 1 , wherein at least one of the third candidate pair or the fourth candidate pair is a same candidate pair as the fifth candidate pair. 
     
     
       5. The method of  claim 1 , wherein at least one of the third candidate pair or the fourth candidate pair is different than the fifth candidate pair. 
     
     
       6. The method of  claim 1 , wherein the first pattern and the second pattern are a same pattern. 
     
     
       7. The method of  claim 1 , wherein the first pattern is different than the second pattern. 
     
     
       8. A method, comprising:
 receiving a list of one or more defect candidates; 
 determining a plurality of candidate pairs from the list to establish a set of candidate pairs, wherein the determining comprises:
 generating a first candidate pair comprising a first defect candidate and a second defect candidate responsive to determining that the first defect candidate and the second defect candidate have a same defect; and 
 generating a second candidate pair comprising a third defect candidate and a fourth defect candidate responsive to determining that the third defect candidate and the fourth defect candidate have a same fault effect; 
 
 dropping the second candidate pair from the set of candidate pairs to generate a reduced set of candidate pairs responsive to determining that the third defect candidate and the fourth defect candidate have the same fault effect; 
 distinguishing a fifth defect candidate of a third candidate pair in the reduced set of candidate pairs from a sixth defect candidate of the third candidate pair using fault inactivation, wherein the fault inactivation comprises forming a first pattern that detects the fifth defect candidate but not the sixth defect candidate; 
 grouping the third candidate pair with a fourth candidate pair in the reduced set of candidate pairs to create a first group responsive to determining that a first portion of a first circuit associated with the third candidate pair does not overlap a second portion of the first circuit associated with the fourth candidate pair; 
 inserting the first group into a second circuit; 
 forming a second pattern for at least some of the first group based upon the second circuit; 
 identifying a convergent point of a third circuit on a first portion of the third circuit that is associated with a seventh defect candidate of a fifth candidate pair and a second portion of the third circuit that is associated with an eighth defect candidate of the fifth candidate pair, wherein a third portion of the third circuit that is associated with the convergent point is merely able to be affected by at least one of the seventh defect candidate or the eighth defect candidate via the convergent point; 
 removing at least some of the third circuit after the convergent point to generate a reduced circuit; and 
 determining whether the seventh defect candidate is equivalent to the eighth defect candidate based upon the reduced circuit. 
 
     
     
       9. The method of  claim 8 , wherein the second circuit and the third circuit are a same circuit. 
     
     
       10. The method of  claim 8 , wherein the second circuit is different than the third circuit. 
     
     
       11. The method of  claim 8 , wherein at least one of the third candidate pair or the fourth candidate pair is a same candidate pair as the fifth candidate pair. 
     
     
       12. The method of  claim 8 , wherein at least one of the third candidate pair or the fourth candidate pair is different than the fifth candidate pair. 
     
     
       13. The method of  claim 8 , wherein the first pattern and the second pattern are a same pattern. 
     
     
       14. The method of  claim 8 , wherein the first pattern is different than the second pattern. 
     
     
       15. A method, comprising:
 receiving a list of one or more defect candidates; 
 determining a plurality of candidate pairs from the list to establish a set of candidate pairs, wherein the determining comprises:
 generating a first candidate pair comprising a first defect candidate and a second defect candidate responsive to determining that the first defect candidate and the second defect candidate have a same defect; and 
 generating a second candidate pair comprising a third defect candidate and a fourth defect candidate responsive to determining that the third defect candidate and the fourth defect candidate have a same fault effect; 
 
 grouping a third candidate pair of the plurality of candidate pairs with a fourth candidate pair of the plurality of candidate pairs to create a first group responsive to determining that a first portion of a first circuit associated with the third candidate pair does not overlap a second portion of the first circuit associated with the fourth candidate pair; 
 inserting the first group into a second circuit; and 
 forming a first pattern for at least some of the first group based upon the second circuit. 
 
     
     
       16. The method of  claim 15 , comprising:
 identifying a convergent point of the second circuit on a first portion of the second circuit that is associated with a fifth defect candidate of the third candidate pair and a second portion of the second circuit that is associated with a sixth defect candidate of the third candidate pair; 
 removing at least some of the second circuit after the convergent point to generate a reduced circuit; and 
 determining whether the fifth defect candidate is equivalent to the sixth defect candidate based upon the reduced circuit. 
 
     
     
       17. The method of  claim 15 , comprising:
 dropping the second candidate pair from the set of candidate pairs to generate a reduced set of candidate pairs responsive to determining that the third defect candidate and the fourth defect candidate have the same fault effect. 
 
     
     
       18. The method of  claim 17 , wherein the third candidate pair and the fourth candidate pair are selected from the reduced set of candidate pairs. 
     
     
       19. The method of  claim 15 , comprising:
 distinguishing a fifth defect candidate of the third candidate pair from a sixth defect candidate of the third candidate pair using fault inactivation prior to the grouping, wherein the distinguishing comprises forming a second pattern that detects the fifth defect candidate but not the sixth defect candidate. 
 
     
     
       20. The method of  claim 19 , wherein the first pattern is different than the second pattern.